small to warrant conclusions. The incidence of anemia was lower in the other intervals and the differences among them were not significant, although the lowest rate observed (23.9 percent) was in the "long" (more than 48 months) group.
The "most striking effect" he observed was the association between toxemia of pregnancy and birth interval. As he stated in his conclusion,
The longer the interval between births, the more likely the mother is to suffer from some form of hypertensive toxemia of pregnancy. The incidence of this complication is lowest when the interval is twelve to twenty-four months, significantly higher when it is twenty-four to forty-eight months, and much higher when it exceeds four years. In the present study this was equally true of white and colored ward patients and private patients. In patients who have had a previous hypertensive toxemia of pregnancy, the likelihood of repetition becomes progressively greater as the interval becomes longer. (62, p. 462)
Eastman was unable in this sample to control for maternal age which is also associated with toxemia of pregnancy. Of course, as interval increases, age of mother increases.
None of the other factors he examined, hemorrhage, infection, or maternal mortality, was found to be associated with birth interval.
THE QUESTION OF CAUSE AND EFFECT
Investigations of the causal role of family size or birth interval in regard to the "effects" described earlier are extremely few in number. The problem, as noted from the start, is complicated by the fact that all of these effects are unquestionably the product of many interacting causal factors. Nevertheless, there is some evidence, both direct and indirect, concerning the place of these two factors in the causal web.
Common Sense Effects
Family Size and Food Expenditures. Wherever families are dependent on cash income for the purchase of food, every additional member of the family adds to the strain on the family food budget. Our study in Candelaria (10) showed this clearly, as may be seen in Figure 13, in which per capita food expenditures per week are plotted against the number of living children per family. In situations in which families must buy their food and when food expenditures fall to such an extent in association with increases in family size, then common sense suggests that nutrition would suffer and the increase in malnutrition with family size that we saw in Table 2 should not be surprising.
There is some evidence to show that this phenomenon is not limited to agricultural day laborers in a developing country. In the United States an